Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Dew -
read INTC’s SEC filings over the past decade and add up the actual cash amounts spent to buy back shares on the open market. You will see that the number far exceeds the cash proceeds INTC has obtained from option exercises during that period.
There you go again. You just don't get it. The options exercised over the last 10 years are for grants that happened anywhere from 5 to 20 years ago. There was a 5 year lag until maturity plus a 5 year period to exercise. So 10 years ago someone could have exercised options that were granted 20 years ago and were coming up on expiration. Today someone could be exercising options granted 5 years ago. So over the 10 year period the exercises could represent grants spanning 15 years.
No matter how hard you try you're just making Intel's case that there's no good way to account for them in the current year.
sgolds -
This is the impression that will start to persist in the world at large - that the QHZ rating loses steam with this processor.
Boy I never expected to be arguing this side but I think AMD is right in perceiving that the customer is dumb, is shopping for MHz and the best strategy is to present a model number that will simply fool the customer into thinking he's buying MHz. AMD should and will continue this marketing scheme for as long as it works and I don't see an end to it anytime soon.
Now should Intel market the 3.0GHz P4 as a P4-3500 and match BS with BS? Maybe so?
Saturn -
Had the stock been sold on the open market, the capital account appreciation would have been greater. So the option is not an expense, but rather a loss of opportunity for the company.
Finally someone makes a valid point. When Intel granted options they set a cap on the value of those shares for the duration of the option period. Yes, they have lost opportunity due to price appreciation but is it right to consider lost opportunity an accounting loss? I regularly write CCs and have on occasion failed to realize all the gain I would have by simply holding the shares. I suffered "opportunity loss". Do you think I can write that off my taxes? On a similar note, many companies write Puts so as to lower the price of the stock buybacks. Does this get accounted for as income or a lowering of the cost basis of the buyback shares?
Haddock -
while Elmer hints that he knows the answer too but can't say
I'm not hinting. I'm telling you I know the number manufactured and it doesn't match up with the number claimed to be sold by the rumor mill. You may not believe me but my point is that I'm not just dropping a subtle hint. What I didn't say was if it was too low or too high.
Dew -
The company receives cash in the amount of the exercise price when an option is exercised. But when the company buys offsetting shares in the open market, it pays the full market price.
I know you just don't get it but I'll try again. The options exercised today are satisfied with shares bought years ago at the then FMV and option strike price. It's a push. The shares bought today are the ones that back the options granted today. Both at FMV. It's a push.
Example:
Company A buys X shares at $Y. Company A grants employees options to buy X shares at price $Y. 5 years from now employees exercise their options to buy X at $Y and company A delivers X at $Y. It's a push. At the same time company A buys X' at $Y' while granting X' options to employees at $Y'. You think the exercised shares are actually the X' shares at $Y. They are not. You are wrong.
Smooth -
But that one little mouse "click" as you put it would cost me a subscription. Something I have not decided on yet.
Greg said the ignore feature doesn't require a subscription.
Dew -
All this was discussed right here a few months ago. You didn’t get it then and, evidently, you still don’t.
I was thinking the same thing about you.
sgolds -
Based on consistent comments I see on benchmark sites quoted here, the new 3200+ is a good overclocker. That means they could have released it at a higher clock rate.
This is a common misconception. Several factors come into play here. First, the device must run at it's rated speed over voltage and temperature to be an official product. Overclockers can control both to a much finer degree than OEM boxes would. Secondly, all devices must have a built in margin to allow for the degradation in performance over time due to hot electron effects in the n-channels and a corresponding effect in the p-channels. All modern processors will degrade over time and all must overclock or they will fail to meet specifications at some point over their lifetime.
Intel chips away at AMD market share
By John G. Spooner
CNET
Advanced Micro Devices lost market share during the first quarter of 2003, despite increasing shipments in PC processors, according to new figures from Mercury Research.
Sunnyvale, Calif.-based AMD had 16.6 per cent of the PC processor market, compared with 18.2 per cent during the first quarter of 2002, giving up 1.6 points of market share on a year-over-year basis. Meanwhile, Intel commanded 81.7 per cent of the PC processor market. That's up nine-tenths of a point from its 2002 first-quarter market share of 80.8 per cent.
But when measured on a sequential basis, AMD had 2.8 points more than it had during the fourth quarter of 2002. Intel, on the other hand, saw a similar sequential decrease, said Dean McCarron, principal analyst at Mercury Research.
AMD fared well on a sequential basis because it was able to begin shipping more processors during the first quarter, after spending the latter half of 2002 sweeping away extra inventories of its Athlon chips.
During the first few months of 2002, when many in the PC industry believed a turnaround was at hand, AMD built up substantial processor inventories. When PC sales did not accelerate the way some thought they would, AMD was left with millions of extra Athlons, causing it to take corrective action, including cutting shipments of new chips. This lowered its market share year over year, according to Mercury Research.
"It was a pretty big bullet to bite," Mr. McCarron said. "The hit [that AMD] took in the third quarter was pretty phenomenal. But that's basically something it had to do if it was going to get its sales to recover."
Mr. McCarron believes that the market will settle down a bit and the two chipmakers' market share figures will be more stable throughout the rest of the year.
"I don't think we'll be seeing three-point moves quarter to quarter for the rest of this year," he said. "It would be surprising to see any movement more than about a point either way."
That doesn't mean that either chipmaker is planning to lose its edge.
Both Intel and AMD are working on faster, higher-performance processors and chipsets, chips that accompany the processors inside PCs. Intel is poised to launch a series of desktop chips - including a new chipset expected next week - that promise a boost in performance for PCs. Later, the company will introduce a faster 3.2 GHz Pentium 4.
Intel helped make way for those chips by cutting prices on some processors over the weekend. It dropped the price on its 2.8 GHz Pentium 4 by $113 (U.S.) or 30 per cent to $262. The company also cut the price of its 2.6 GHz and 2.66 GHz Pentium 4s by $48 or 20 per cent, to $193.
AMD is working on bringing out higher-performance silicon. The chipmaker is prepping a new Athlon XP 3200+ chip for desktops and is on track to launch its next-generation chip, the Athlon 64, in September, company representatives have said.
"Going forward, the question will be, 'How does Athlon 64 ramp up?' It could well be that the Athlon XP chip becomes AMD's new value line and Athlon 64 becomes the performance line," Mr. McCarron said. "Given where Athlon XP is priced, the low end of the line is essentially against where Intel is with Celeron."
Mercury Research's overall figures include unit shipments of PC processors but don't account for whether or not the chips are built into a PC. Also included in Mercury's figures are the chips that Intel sells to Microsoft for its Xbox video game console. Without Xbox, which runs on an older Intel processor, Intel held 80.9 per cent of the PC processor market during the first quarter of 2003 and AMD had 17.3 percent of the market, Mr. McCarron said.
About 1.7 per cent of sales belonged to the "other" category, which includes chipmakers Via Technologies and Transmeta.
Carl -
Intel is (and has been) spending 1 Billion from earnings each quarter buying back shares on the open market. This has been going on for the last couple of years, I believe, if the sources explaining this "trick" are correct.
Those repurchased shares are for the options granted this year. They are purchased at fair market value (FMV) and the options are granted at FMV. It's a push. The options exercised today represent the shares purchased years ago at FMV and granted at FMV. It's a push.
Haddock -
I wonder why Intel is so secretive about this. I can't help thinking the real figures must range below the level of 'mindblowing'. Lets say they sold a pile of them to the NSA, they could still tell us about the rest of them, couldn't they?
I only said there was a discrepency. I didn't say if it was more or less...
doug -
Oh come on, now. Sure, a 2.3 GHz core would have been more comfortable for the 3200+, but I think it's about time AMD stood up to Intel's benchmark deception. Intel can continue to buy off this or that benchmark or website (Tom's), and AMD can continue to say, sorry, we're not going to play by your benchmarks-- we'll label it as we feel appropriate, and let the market be the judge.
And you can continue to make up conspiracy theories to explain away AMD's fraudulent marketing schemes. Bottom line is that AMD relies on customers mistaking the phoney QS number for actual GHz. They don't care how it performs, they just roll out a new QS number on schedule regardless of the performance.
Not to get into the whole QS thing again, but if anything demonstrates its success, it's the reaction the Intel fans are having with every new speed grade.
Not me. I'm an AMD investor and I'm happy to make money based on AMD deception. It'll all come out in the wash.
wbmw -
Speaking of which, do you know anyone here who would like to buy some calls from me? I am willing to part with June $10.00's for $0.50
The bid is only $0.05 for those June $10 calls. I sold some of those myself a couple of weeks ago. I fully expect them to expire worthless.
By the way, where's Niceguy when you need him? He was always willing to give up another pint of blood, etc.
Yes I remember him!!! That guy has to be the biggest loser of all time, and I'm speaking about traders here not personalities. For all we know he's here posting under another alias, just like TheWatsonYouth.
BTW, TWY started posting here under the name "Andy Grove". I PM'd the sysadmin and complained that it was wrong for someone to claim to be the Chairman of the Board of a major corporation related to the subject matter here and now you see him as "Andy Grave". Same old TWY.
wbmw -
most of our benchmarks indicate that the 3200+ only offers incremental improvements over the 3000+. In many cases, it simply can't keep up with the Pentium 4 3.0GHz running with an 800MHz (effective) FSB.
You can't really blame AMD for continuing this line of deception. As long as the public lets them get away with it they're going to continue doing it.
You should ask yourself, am I an Investor or an Intel advocate? If you find that you are both then go with the flow. AMD doesn't have to be credible to the technically savy to make you money. If the public is stupid enough to believe AMD's QS numbers then you're not going to turn the tide here. You know it's BS, I know it's BS and even the AMD fans here know it's BS but we don't make the rules. If we set our personal feelings aside we can profit from the baloney.
Edgar -
You and I both know somebody at Intel knew about this and never released the info to builders.
No We don't know that. You want to believe it but it's a less than stellar notion. You hate Intel and want to believe the most outrageous conspiracy you can come up with.
Notice the problem, if they kick the cpu speed down it runs fine. Hmmm I wonder what that does to their benches now.
It obviously lowers the scores on the small percentage of units this represents. The ones that run properly at full speed still have the same scores.
If I remember right wasn't all the big news a little bit ago about how Intel was going to sell 25,0000 Itaniums this year alone.
Could be. Lots of people who have no idea what Intel's real volumes are like to publish numbers they can't substantiate.
I would like to add that I know what Intel's manufacturing volumes are for Itanium 2 and my ability to comment is somewhat restricted. I will say that the number manufactured does not line up well with the number reported by non Intel sources as being sold. I find the disparity hard to reconcile.
wbmw -
Even so, with issues like the bug reported this morning, Intel has to maintain a strong message and some very compelling proof points.
Immediate identification of the problem and a prompt solution, including free replacement, goes a long way to establishing a "strong message". How long did it take SUNW to identify the cache problems they had a couple of years ago? Remember when they required their customers to sign a non disclosure agreement just to get the fix?
wbmw -
I am glad you are adamant about Carl's free speech here, but in this case, even I agree that his posts have little content worth viewing, and this can be especially annoying to those without the filter option.
If they have no "ignore" option then I understand their annoyance. I didn't realize they had no "ignore". Funny they didn't mention it.
EP
Edgar -
The Itanium debacle is what Opteron needs...
Debacle??? You wish... You guys are getting a little carried away here. Issue identified, solution in place.
chipguy -
Intel made more than that during sampling and it has been in full production since last July.
I don't know how you know that and I won't ask. An important point is that it's possible the volume estimates are off base and perhaps by a wide margin.
EP
Smooth -
Oh stop.
What do you mean stop?
You have the power to easily stop viewing all posts from Carl. For some reason that's not good enough for you. You want to silence him so nobody else can read his posts.
Just ignore him and you'll never be bothered by his posts again. One tiny little mouse click and the problem is gone.
Smooth2o -
Sorry. Someone who cannot debate the issue and rants, is only an annoyance.
Then ignore him. Do you have something against the rest of us reading his posts?
Haddock -
I don't see how that would have prevented a PR coup.
I don't supposed it would have but it would have been to a limited audience.
Smooth2o
I suggest you complain to the moderator. I am.
Carl's posts are about Intel and policy that affects the share price. They're on topic if somewhat annoying. If you don't like reading them then I suggest you use the ignore feature. It's easy to do.
Haddock -
AMD would have had the fastest SPECint2000peak figure in the world for a moment. That would have been a PR coup, no doubt about it.
I doubt it. Those who follow SPEC know to use the base number.
Haddock -
Sounds like a missing speed path in the test coverage, since the problem is not present at 800MHz. In fact it sounds rather like the P3-1.13GHz problem.
Yes it does but in fairness, there are more speed paths than can possibly be tested. There is no know method of quanifying speedpath coverage at test. It is an empirical process.
Carl -
Please explain what CEO of Schwab, Mr. Pottruck gets paid for his job on the board of Intel. $1000 per meeting?
Why would anyone take a job that doesn't pay?
wbmw -
Let me know when you find someone else who can back up AMD's assertion that their 3000+ Athlon performs 17% faster than a 3.06GHz Pentium 4 with Hyperthreading.
If Dan3 is still around I'm sure He'll vouch for it.
Keith -
While it´s true that the new P4 3Ghz was released prematurely to avoid AMD´s claims of highest SPEC performance at launch
If it's true then you should be able to substantiate this statement. Otherwise it would appear you are making it up.
Another note: Intel holds the highest SPEC scores without the 3.0GHz P4. The 3.06GHz beats Opteron as well (1099 vrs 1095) in SPECint so your statement is looking more and more like pure nonsense.
Tenchu -
Can you tell me what "QuantiSpeed" is and why AMD pretended like it's a radical new feature of Athlon XP?
I can.
Realizing customers are shopping with frequency as one of their decision criteria, AMD invented "QuantiSpeed" with the specific intent of tricking the customer into believing the "QS" number was actually the frequency number the customer was looking for. The QS marketing scheme relies entirely on the customer being fooled into selecting the AMD machine with the QS rating he has mistaken for actual frequency.
While some may argue that the customer will in the end get more or less the performance he was looking for, and therefore no actual fraud was committed, the scheme nevertheless relies on the belief that the customer is too stupid to make the right decision and therefore must be tricked into selecting the AMD machine that is best for him.
TWY -
It is not based on inside information. It only serves to illustrate the number of good chips possible assuming 2500W/ week and 100mm^2 chips and 50% yield.
Yes it shows the number of good die from a process with unacceptably high defect density. It is unlikely that you would use your best guess about wafer capacity and then not use your best guess about defect density, but who knows. It is actually close to the number I would have "estimated" for an SOI process.
TWY -
As a refurbished 12 inch fab, I would guess it is capable of perhaps 2500 wafers per week in a strict manufacturing mode. Remember though, we would be talking 65nm node so that still means about 10 million chips per quarter assuming 100mm^2 chips and 50% yields.
You are assuming pretty bad defect density in your calculations. I will assume this is based on inside information....
Tenchu -
Ahhh my old buddy Fuchi Wu.... Yes those were the days....
YB -
Elmer, c'mon - how many desktop Opterons the market can eat? Maybe half million or so per quarter with prices $300 to $600. First, it's not that much that the huge die will affect the overall AMD fab output. Second, it's good money. It's $200 mil per quarter.
You're missing my point but it's not worth the trouble.
UpNDown -
Now, that's ridiculous. 1) Opteron does not need a 2-way system to set itself apart from P4/Xeon.
You are over reacting here. My point was that Opteron is too big to compete on the desktop where it doesn't have the distinguishing performance lead that it does in a 2-way system. If AMD wants to compete with Opteron and it's huge die then it's fine with me.
Edgar
Advanced seems to be firing on all cylinders. Barton 3200+, Opteron, Flash & upcoming Athlon 64.
Multi- hundred million $$ losses, dwindling cash reserves, debt downgrades, looming bankruptcy... Ahhhh it just doesn't get any better than this!
YB -
think July $5 options are cheap.
Hey I like that but I wouldn't hold them through earnings if I were you.
I hope this message doesn't get deleted...
Dew -
The survey results might be more revealing if those on the message board who are Intel employees refrained from voting.
Who are those people?
sgolds -
Elmer, different market. Athlon64 is for a competitively prices 64-bit workstation or laptop. While there may be no need for a high end 1P Opteron workstation, there is always a need for 1P machines for the masses!
That's the point I was getting at. If Opteron needs a 2-way system to set ifself apart from P4/Xeon, and with A64 not looking particularly strong, how will A64 fare on the desktop?
I hope this post doesn't get deleted.
sgolds -
yourbankruptcy, upndown, you both make good arguments that the incremental cost of a 2P system may make a 1P system undesirable. That does not negate my speculation that a 1P system would be a ho-hum product because you don't get the advantages of aHT. Both reasons seem to reinforce each other.
You all have made a good case for the lack of need for a uniprocessor workstation but what does this bode for Athlon64?
Can AMD return to profitability on Opteron alone?
EP
I hope this posts doesn't get deleted too...
sgolds -
Elmer, yes, the single CPU version should be the easiest. Then why wasn't it released last month?
Which brings us back to the original point. None were released last month. So the question remains, can they produce them?